This document summarizes research into issues that occur with skewed bridge construction in Ohio. Line girder analysis does not accurately predict behavior due to intermediate and end crossframe effects. A refined analysis including crossframe interaction showed more uniform differential deflection. Parametric studies found line girder analysis can conservatively estimate forces for skews up to 45 degrees. Detailing methods like full dead load fitting or lean-on bracing can minimize twisting issues during deck placement.
This document summarizes research into issues that occur with skewed bridge construction in Ohio. Line girder analysis does not accurately predict behavior due to intermediate and end crossframe effects. A refined analysis including crossframe interaction showed more uniform differential deflection. Parametric studies found line girder analysis can conservatively estimate forces for skews up to 45 degrees. Detailing methods like full dead load fitting or lean-on bracing can minimize twisting issues during deck placement.
This document summarizes research into issues that occur with skewed bridge construction in Ohio. Line girder analysis does not accurately predict behavior due to intermediate and end crossframe effects. A refined analysis including crossframe interaction showed more uniform differential deflection. Parametric studies found line girder analysis can conservatively estimate forces for skews up to 45 degrees. Detailing methods like full dead load fitting or lean-on bracing can minimize twisting issues during deck placement.
This document summarizes research into issues that occur with skewed bridge construction in Ohio. Line girder analysis does not accurately predict behavior due to intermediate and end crossframe effects. A refined analysis including crossframe interaction showed more uniform differential deflection. Parametric studies found line girder analysis can conservatively estimate forces for skews up to 45 degrees. Detailing methods like full dead load fitting or lean-on bracing can minimize twisting issues during deck placement.
Skewed Bridges Travis Butz, PE Burgess & Niple, Inc.
Ohios Skew Problems:
Recurring constructability problems during deck pours Predicted deflections disagree with field results
Decks with exposed rebar, poor finish, inconsistent
thickness Excessive girder twist in one case, capacity of the structure was compromised A study was commissioned to identify causes and to recommend solutions - Why is this happening? - What analysis methods are appropriate? - How can we prevent these problems?
Skewed Bridge Behavior
Out-of-plane effects occur in skewed bridges that cannot be predicted by line girder analysis methods (neglecting crossframe effects). AASHTO/NSBA Guidelines for Design for Constructability identifies two separate issues:
Intermediate Crossframe Effects
End Crossframe Effects
Test Case, Intermediate Crossframe Effects:
FRAMING PLAN
TRANSVERSE SECTION
Test Structure, Deflection Due to Deck Weight
Line Girder Analysis Results Crossframe Effects Ignored Results Show: Large differential deflections between interior and exterior girders
Crossframe Locations
5 Deflection (in)
Abrupt changes in differential deflection across the width of the bridge
4 3 2 G5
G4
G3
G2
G1
1 0 0.00
50.00
100.00
Length (ft)
150.00
200.00
Line Girder Analysis Results Crossframe Effects Ignored
vertical, the crossframes will not permit differential deflections of this magnitude. Conclusion: Crossframe interaction needs to be included to accurately model structure behavior.
Lengthened
Shortened
Differential vertical deflection causes crossframes to
deform if the girders do not twist. Large forces are needed to create axial deformations in the crossframe members, so resistance to this type of deflection is very high.
Undeformed
Undeformed
Twisting of the girders allows differential deflection to
occur without deforming the crossframe. The torsional stiffness of the girders is low compared to the stiffness of the crossframes, so this behavior is dominant.
Test Structure, Deflection Due to Deck Weight
Refined Analysis Results Intermediate Crossframe Effects Included Results Show:
Crossframe Locations
5 Deflection (in)
More uniform differential deflection across the width of the bridge at crossframe locations (compared to line girder analysis)
4 3 2 G5
G4
G3
G2
G1
1 0 0.00
50.00
100.00 Length (ft)
150.00
200.00
Line Girder Analysis Results Crossframe Effects Ignored
Differential Deflection (in)
Section D-D
Refined Analysis Results Intermediate Crossframe Effects Included
Girder Deflection (in)
Differential Deflection (in)
Girder Deflection (in)
Section D-D
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
A
A Differential Vertical Deflection (inches)
Section A-A
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
B
B Differential Vertical Deflection (inches)
Section B-B
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
C
C Differential Vertical Deflection (inches)
Section C-C
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
D
D Differential Vertical Deflection (inches)
Section D-D
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
E
E Differential Vertical Deflection (inches)
Section E-E
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
F
F Differential Vertical Deflection (inches)
Section F-F
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
G
G Differential Vertical Deflection (inches)
Section G-G
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
H
H Differential Vertical Deflection (inches)
Section H-H
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
J
Differential Vertical Deflection (inches)
Section J-J
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
K
Differential Vertical Deflection (inches)
Section K-K
Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
L
L Differential Vertical Deflection (inches)
Section L-L Deflections Exaggerated x 12
Deflected Shape Due to Intermediate Crossframe Effects (Refined Analysis):
M
M Differential Vertical Deflection (inches)
Section M-M
Deflections Exaggerated x 12
Support Reactions Due to Wet Concrete Weight, Refined Analysis:
Forward Bearings (Exp.) Rear Bearings (Fixed)
Rear Bearings (Fixed)
Forward Bearings (Exp.)
End Crossframe Effects
To illustrate end crossframe behavior, we will examine a 2girder structure with end crossframes only (no intermediate bracing).
This illustration is adapted from Beckmann & Medlock, 2005
PLAN VIEW
2-Girder Structure:
ISOMETRIC VIEW (PARTIAL)
The end diaphragm can be thought of as a pair of rigid
links connecting the top flange of one girder to the bottom flange of the adjacent girder.
Deflection of a Cambered Girder:
When a girder deflects, the top flange moves
longitudinally relative to the bottom flange at the beam ends. We will define this distance as .
2-Girder Structure:
GIRDER A
GIRDER B
Dx PLAN VIEW (PARTIAL)
The end crossframe of a skewed structure restrains
the longitudinal translation of the top flange.
2-Girder Structure:
GIRDER A
Dy
GIRDER B
Dx PLAN VIEW (PARTIAL)
The end crossframe forces the top flange to move
radially about the adjacent bearing point. The resulting motion produces twist in the girders.
2-Girder Structure:
Dx
GIRDER B
Dy
Dy
GIRDER A
Dx
PLAN VIEW (PARTIAL)
The movement of the top flange is approximately
perpendicular to the centerline of bearings.
Test Structure, Girder End Twist
End Crossframes Only:
Forward
Sign Convention: (+ Clockwise, Looking Forward
- Counterclockwise, Looking Forward)
Intermediate Crossframes Only:
Forward
Sign Convention: (+ Clockwise, Looking Forward
- Counterclockwise, Looking Forward)
Combined Effects: Girder End Twist
End Crossframes Only / Intermediate Crossframes Only:
Forward
Sign Convention: (+ Clockwise, Looking Forward
- Counterclockwise, Looking Forward)
Combined effects: Forward
Sign Convention: (+ Clockwise, Looking Forward
- Counterclockwise, Looking Forward)
Evaluation of Analysis Methods
Parametric Study:
Parametric Study, Analysis Methods:
Girder modeled using beam elements
Mz
Line Girder Analysis
Parametric Study, Analysis Methods:
Study Conclusions: When the effects of intermediate crossframes are considered, significant redistribution of shear and moment occurs across the width of the structure. For the structures studied, line girder analysis can be used to conservatively calculate member forces for skews up to 45.
2D Grid Analysis vs. 2D Grid Analysis w/ Truss Crossframes:
Little variation was observed between the girder and intermediate crossframe forces obtained from 2D grid analysis with truss crossframes. The use of 2D grid analysis was shown to be generally accurate in the calculation of moments and shears for the cases investigated. Note that higher levels of analysis provide more precise results, and are recommended when higher precision is needed, or with more complex structures (variable skews, partial length girders, etc.).
2D Grid Analysis w/ Truss Crossframes vs. 3D FEM:
The moment and shear results obtained from 3D FEM analysis show general agreement with the results obtained from 2D grid analysis with truss crossframes.. Although the 2D grid was found to be generally accurate for calculating moments and shears for the structures investigated, 3D FEM analysis does provide more precise results. 3D FEM is recommended for more complex structures.
Detailing Methods Method 1 Steel dead load fit members are detailed to fit with webs plumb with steel dead load on the structure, but not the deck load
Erect girders plumb
Install crossframes Girders rotate out of plumb during deck placement Girders will be permanently twisted
Question: How much twist is acceptable?
Detailing Methods Method 2 Full dead load fit members are detailed to fit with webs plumb with full non-composite dead load of steel and concrete. Erect girders out-of-plumb Install crossframes Girders rotate to vertical during deck placement Girders webs will be vertical in the finished structure
Detailing Methods Method 2 Full dead load fit members are detailed to fit with webs plumb with full non-composite dead load of steel and concrete. This method is generally recommended for skewed bridges by industry experts.
ODOT is not comfortable with erecting girders in an out-of-plumb
position. Steel Dead Load fit is required by ODOT policy.
Detailing Methods Method 3 Lean-on Bracing Use of an alternative lateral bracing system to minimize or eliminate intermediate crossframe effects. Some crossframes are replaced with top and bottom struts only during the deck pour
Lean-on braces allow differential vertical deflection to occur
between girders without inducing twist.
Lean-on Bracing Two types: Internal and External
In an Internal Lean-on System, bracing is provided by a
crossframe located within the portion of the structure that is being loaded.
Lean-on Bracing Internal Lean-on System In an internal system, crossframe locations can be selected strategically to minimize twist in the system. Designers must perform calculations to ensure adequate strength and stiffness are provided. A
X X
Section A-A
Lean-on Bracing External Lean-on System
In an External Lean-on System, the structure is braced
against an external support or a portion of the structure that will not be loaded during the deck pour.
Can girder twist be calculated using line girder results?
For low skews, girder twist can be
estimated using line girder analysis. From AASHTO/NSBA Steel Bridge Erection Guide Specification, erection tolerance = 1/8 per foot of web depth Data shows this method to be conservative up to a 45 skew.
30 < Skew 45
Perform Line Girder
Analysis
Differential Deflections < S/100
ODOT Policy: Skewed Bridge Design Process No
Stiffen Design: 0% to 25% Additional Steel
No
Stiffen Design: No 0% to 25% Additional Steel
Girder Twist < 1/8/ft?
Yes Check That Design Rates Using PC-BARS
Implement External Lean-on Bracing
No
Perform Refined Analysis
Yes
Yes Design Using Line Girder Analysis
Differential Deflections < S/100
Skew > 45
Yes
Implement Internal Lean-on Bracing with Refined Analysis
No
Girder Twist < 1/8/ft?
Girder Twist < 1/8/ft?
Finish Design Using Refined Analysis: Erect Girders Vertical And Allow To Rotate
Finish Design Using Refined Analysis: Erect Girders Vertical And Allow To Rotate
Check That Design
Rates Using PC-BARS
30 < Skew 45
Perform Line Girder
Analysis
Differential Deflections < S/100
ODOT Policy: Skewed Bridge Design Process No
No
Stiffen Design: 0% to 25% Additional Steel
Differential Deflections < S/100
Stiffen Design: No 0% to 25% Additional Steel
Girder Twist < 1/8/ft?
Implement Internal Lean-on Bracing with Refined Analysis
Girder Twist < 1/8/ft?
No
Perform Refined Analysis
Girder Twist < 1/8/ft?
Finish Design Using Refined Analysis: Erect Girders Vertical And Allow To Rotate
Yes
Check That Design
Rates Using PC-BARS
30 < Skew 45
Perform Line Girder
Analysis
Differential Deflections < S/100
ODOT Policy: Skewed Bridge Design Process No
No
Check That Design
Rates Using PC-BARS
Implement External Lean-on Bracing
Stiffen Design: 0% to 25% Additional Steel
Girder Twist < 1/8/ft?
Implement Internal Lean-on Bracing with Refined Analysis
No
Girder Twist < 1/8/ft?
Differential Deflections < S/100
Stiffen Design: 0% to 25% Additional Steel
No
Perform Refined Analysis
Girder Twist < 1/8/ft?
30 < Skew 45
Perform Line Girder
Analysis
Differential Deflections < S/100
ODOT Policy: Skewed Bridge Design Process No
Stiffen Design: 0% to 25% Additional Steel
No
Stiffen Design: No 0% to 25% Additional Steel
Girder Twist < 1/8/ft?
Yes Check That Design Rates Using PC-BARS
Implement External Lean-on Bracing
No
Perform Refined Analysis
Yes
Yes Design Using Line Girder Analysis
Differential Deflections < S/100
Skew > 45
Yes
Implement Internal Lean-on Bracing with Refined Analysis
No
Girder Twist < 1/8/ft?
Girder Twist < 1/8/ft?
Finish Design Using Refined Analysis: Erect Girders Vertical And Allow To Rotate
Yes
Check That Design
Rates Using PC-BARS
ODOT Policy: End Crossframes
For skews > 30 degrees, do not install end crossframe diagonals until deck placement in the adjacent span is complete End Armor
Condition at Deck Placement: Note that the girder ends are unbraced. Temporary bracing may be required.
End Armor
Final Condition: Diagonals installed
Bottom Chord Diagonals
Bottom Chord
Refined Analysis using Midas Civil:
In midas civil user can model the construction sequence considering the girder lift, installation and the deck pouring sequence. The shell elements works well in determination of the girder twist. A study has been done in midas civil for the determination of the girder twist during the deck pouring. The following pouring sequence has been assumed:
Stage 1
Stage 3
Stage 2
The following Stages were modeled:
1. Stage 1: Steel Girders are installed and self weight of steel is activated. 2. Stage 2: The scaffolding load is activated. The load is activated in the following fashion for the overhangs:
3. Stage 3: The deck dead load is activated for the deck pour 1. 4. Stage 4: The deck dead load is activated for the deck pour 2.
5. Stage 5: The deck dead load is activated for the deck pour 3.
Deformation Results: Twisting during deck 1 pouring
Deformation Results: Twisting during deck 2 pouring
Deformation Results: Twisting during deck 3 pouring
=> Twisting can be accurately
estimated by Midas Civil so that proper measures can be taken
